You are discussing how you evaluate machine learning models after training and before release. The interviewer wants to understand how you think about choosing metrics, reading tradeoffs, and deciding whether a model is good enough for use.
What is your experience with machine learning model evaluation techniques?
Choosing the right evaluation metric for the business problemInterpreting precision, recall, and F1 togetherUsing confusion matrix counts to explain model behaviorAdjusting thresholds based on operational tradeoffs